Skip to main content
Version: Python

pct

agg.pct returns an aggregator that computes the designated percentile of values, within an aggregation group, for each input column.

Syntax​

pct(percentile: float, cols: Union[str, list[str]], average_evenly_divided = False) -> Aggregation

Parameters​

ParameterTypeDescription
percentilefloat

The percentile to calculate.

colsUnion[str, list[str]]

The source column(s) for the calculations.

  • ["X"] will output the designated percentile value in the X column for each group.
  • ["Y = X"] will output the designated percentile value in the X column for each group and rename it to Y.
  • ["X, A = B"] will output the designated percentile value in the X column for each group and the designated percentile value in the B value column renaming it to A.
average_evenly_dividedbool

When the percentile splits the group into two halves, whether to average the two middle values for the output value.

  • When set to True, the two middle values are averaged.
  • When set to False, the smaller value is used. The default is False. This flag is only valid for numeric types.
caution

If an aggregation does not rename the resulting column, the aggregation column will appear in the output table, not the input column. If multiple aggregations on the same column do not rename the resulting columns, an error will result, because the aggregations are trying to create multiple columns with the same name. For example, in table.agg_by([agg.sum_(cols=[“X”]), agg.avg(cols=["X"]), both the sum and the average aggregators produce column X, which results in an error.

Returns​

An aggregator that computes the designated percentile value, within an aggregation group, for each input column.

Examples​

In this example, agg.pct returns the 68th percentile value Number as grouped by X.

from deephaven import new_table
from deephaven.column import string_col, int_col, double_col
from deephaven import agg as agg

source = new_table(
    [
        string_col("X", ["A", "B", "A", "C", "B", "A", "B", "B", "C"]),
        string_col("Y", ["M", "N", "O", "N", "P", "M", "O", "P", "M"]),
        int_col("Number", [55, 76, 20, 130, 230, 50, 73, 137, 214]),
    ]
)

result = source.agg_by(
    [agg.pct(percentile=0.68, cols=["PctNumber = Number"])], by=["X"]
)

In this example, agg.pct returns the 68th percentile value Number and the 99th percentile value Number as grouped by X.

from deephaven import new_table
from deephaven.column import string_col, int_col, double_col
from deephaven import agg as agg

source = new_table(
    [
        string_col("X", ["A", "B", "A", "C", "B", "A", "B", "B", "C"]),
        string_col("Y", ["M", "N", "O", "N", "P", "M", "O", "P", "M"]),
        int_col("Number", [55, 76, 20, 130, 230, 50, 73, 137, 214]),
    ]
)

result = source.agg_by(
    [
        agg.pct(percentile=0.68, cols=["Pct68Number = Number"]),
        agg.pct(percentile=0.99, cols=["Pct99Number = Number"]),
    ],
    by=["X"],
)

In this example, agg.pct returns the 97th percentile value Number, and agg.median returns the median Number, as grouped by X. A second result table is then created to show the difference in output when average_evenly_divided is set to True.

from deephaven import new_table
from deephaven.column import string_col, int_col, double_col
from deephaven import agg as agg

source = new_table(
    [
        string_col("X", ["A", "B", "A", "C", "B", "A", "B", "B", "C"]),
        string_col("Y", ["M", "P", "O", "N", "P", "M", "O", "P", "N"]),
        int_col("Number", [55, 76, 20, 130, 230, 50, 73, 137, 214]),
    ]
)

result = source.agg_by(
    [
        agg.pct(percentile=0.97, cols=["Pct97Number = Number"]),
        agg.median(cols=["MedNumber = Number"]),
    ],
    by=["X"],
)
result = source.agg_by(
    [
        agg.pct(
            percentile=0.97, cols=["Pct97Number = Number"], average_evenly_divided=True
        ),
        agg.median(cols=["MedNumber = Number"]),
    ],
    by=["X"],
)